Hitchhikers Guide Part 4

And some early thoughts on the state of pre-season reactions

Before diving into the next group of fragile starters we have on tap for today’s edition of the Hitchhiker’s guide, I want to address everyone’s favourite August tradition: pre-season usage narrative crafting.

If you were to talk to a person who had never played fantasy football before, and tried to explain the phenomenon of pre-season fantasy twitter you would draw faces blanker than those adorning Neutral Milk Hotel’s In the Aeroplane Over the Sea.

---

“Well Jeff, you see this is the first pre-season game… So actually you want players to not play… And if they do play, you need them to leave the game as soon as possible… or at least before the rest of the first tier backups leave… But also we need to track which players play in which personnel groupings… because if WRs only play in 11 personnel then we can’t draft them… but Tight Ends need to play in 11 personnel or we can’t draft them… and this team didn’t play a 3rd down so we really don’t know if this running back plays LDD or not… LDD means long down and distance by the way, it’s not a new psychedelic drug.”

“Ok, but did the player make any plays?” asks Jeff.

“I have not the slightest idea,” you answer, instead feverishly gesturing toward a Nathan Jahnke chart you inscribed upon a stone tablet.

---

My Annual Pre-Season Rant

I say most of the above in jest. Some of this information is legitimately useful. The issue is that we are starved for new information that we likely over-estimate just how useful. At best, pre-season usage offers *some* indication of how the team would operate in neutral-script if Week 1 started today.

But we know teams don’t play the entire game in neutral script. And we certainly know usage is not static throughout the whole season. We are also trying to extrapolate information from tiny samples in exhibition games, with no certainty as to the actual motivation for each choice a coach makes. Even to the extent the takeaways we have are useful, we are at continual risk of simply being wrong.

For much of the off-season, we don’t have information about the ‘micro’ aspects of each player’s case, and thus we tend to focus our energy on the ‘macro’ aspects: zooming out from a player’s projected routes and snaps and into broader analysis of their talent profile, team environment and range of outcomes.

Combining the macro with an estimation of the micro is exactly what we try to do in this column. But it is crucial to treat each micro-input we get as a new puzzle piece to combine with our existing priors rather than the answers to the test.

For instance, was it part of Rondale Moore’s range of outcomes that he be the planned starter in 2-WR sets to open the season? Yes. Does the fact he played only in the slot for his (five snap!) Week 1 pre-season sample hurt the chances that scenario plays out? Yes. Do I now want to draft Michael Wilson before him? Absolutely not.

We aren’t drafting snaps, routes and carries. We are drafting players. We want to keep the focus on the entirety of each player’s individual profile, from college to the NFL, and blend that with our even further zoomed out analysis of that player’s archetype. (Rookie, Y2, ADP market trends etc.)

These help paint a picture of what to expect from that player contingent on several usage scenarios that may or may not be stable across a full season. The players we are most interested in — especially when discussing the middle and late rounds — are the talent-focused bets we feel strongly about the upside of *IF* the type of usage we need manifests at some point over the season. Recall of course that usage and talent are inherently correlated. Part of the reason we are trying to make “anti-fragile” talent bets, is to target players who will dictate increased usage through the strength of their play. This was the focus of my argument on twitter today regarding Travis Etienne. Coaches and GMs have to build rosters with a range of outcomes in mind for each player’s health and skill development in mind, and we have to as well. Not every plan is set in stone.

This isn’t to say I straight up don’t care about pre-season usage. I absolutely do. But given the tendency of the fantasy market to over-react to small samples encapsulating a moment in time, most legitimate shifts are likely to be baked further into players’ costs than I think is appropriate.

One note: an area the fantasy community (at least based on my twitter feed) has seemingly improved on is its assessment of rookies in pre-season. Despite starting the pre-season “fifth", “fourth” and “third” on the depth chart respectively, I saw primarily praise for Roschon Johnson, De’Von Achane and Zach Charbonnet on the timeline. Every coach treats rookies differently and their usage is especially prone to dramatic shifts during pre-season and the early season.

We should care much more about how these players are used (Achane was used on a Texas route as predicted!), and how they play, than what order they enter the game. Recall, Garrett Wilson was the WR4 to open pre-season and suffered a legitimate ADP fall last year as a result. Nonetheless, he ran as the WR3 come week 1, and then showed why we bet on talent over role for anti-fragile player profiles: he quickly demonstrated to the Jets that they should never take him off the field.

The main questions I ask to determine how much I care about pre-season are:

1. Is this usage explainable by anything equally intuitive as it being a comment on that player’s current standing?

   1. I.E. rookies being brought along slowly / veteran deference ; player managing an injury ; player playing new position / new team etc.

      • For example: D’Andre Swift playing while Gainwell sat is not a major concern to me because they rested both incumbent backs while playing all the new additions. Swift starting, leaving after six snaps, and breaking an impressive run has me feeling (slightly) better about him now than prior.

2. Is there a larger sample of data on this player with this staff being used counter to the pre-season data that we think is more reliable?

   1. I.E. Christian Kirk playing in 11P only in the pre-season after being used much more frequently than Zay Jones in 12P in 2022: I would still bet on Kirk to lead Zay in routes over the full season.

3. How much of the case for drafting this player is based on their Week 1 and/or neutral script role?

4. Is this usage in line with reporting out of camp regarding that player or counter to that reporting?

   1. For example, Greg Dulcich has drawn rave reviews throughout camp as a pass catcher. Therefore, while his limited pre-season usage (7 routes on 14 drop backs, played only on third downs and 12P) is concerning, I expect Denver to still prioritize him as a pass-catching option over a full game and full season due to both his profile and his reported level of play in camp.

   2. Counter to this example is Rashaad Penny’s questionable usage (27 snaps in a pre-season game including work with the 3rd team, played behind Swift while Gainwell and Scott sat out). He’s drawn relatively middling reviews and was reported to have the most carries with the third team in camp this week. This usage fits in with the larger reporting that suggests he may be behind Gainwell and Swift and I am legitimately shifting my expectations for Penny’s early-season role downward since I last wrote about that backfield.

5. Is any appropriate shift priced in by the market?

   1. This is the biggest aspect to all this — Greg Dulcich being a (likely) perfect example. I do think there is legitimate reason to downgrade him based on the pre-season usage, and he will move down in my ranks next update. But based on the tenor of conversation around him on twitter, I suspect I will move him down much less than the market. Given I was already above the market on him, it’s possible I wind up drafting a higher percentage of Dulcich now than I was prior.

With all this out of the way, let’s continue the Hitchhiker’s Guide to talk some RBs!

---

Hitchhiker’s Guide Part 4

This continues the portion of the HHG dedicated to “fragile backfields.” For more on what those are, and how we approach them in drafts generally, make sure to read the last edition of the HHG if you haven’t. We dove into drafting RBs in the mid rounds at a more conceptual and took a deeper look at the Vikings and Panthers.

NOTE: all data is from the FantasyPoints Data Suite unless otherwise sourced with exception of the following: Rush Yards over Expected (RYOE) and Percentage of Rushes over Expectation (ROE%) are sourced from NFL Next Gen Stats. PFF Grades and Yards per Route Run are from PFF. BAE Rating and Relative Success Rate (RSR) are from Noah Hills. Juke Rate is from PlayerProfiler.

SEE: “Metrics Legend” at the bottom for an explanation of each stat and its acronym

Today we will cover two more fragile starters in Cam Akers and Dameon Pierce, and comment on their backfield mates as both threats to the starter and late-round options.

Cam Akers

UNDERDOG ADP: 66.5 — RB20 (August 7th)

Talent Profile

Cam Akers may be the most difficult RB to analyze from a talent perspective in the NFL. We can effectively divide his career into four parts. Trying to thread those parts together into one single takeaway as to who Akers is and has been is nearly impossible, but we will try our best to sketch out the possibilities.